A retrospective evaluation was performed of the survival after conservative therapy of infants with persistent pulmonary hypertension who met the published criteria of Bartlett et al (Pediatrics. 1985;76:479-487) or Short et al (Clinics in Perinatology. 1987;14:737-748) for extracorporeal membrane oxygenation (ECMO) therapy. An 80% to 90% mortality rate can be predicted with these criteria, which are based on historical data, if ECMO is not used. The records of infants with the diagnosis of persistent pulmonary hypertension, weighing greater than 2 kg at birth and who were treated during two time periods, January 1980 to December 1981 [23 patients] and January 1986 to December 1988 [17 patients], were reviewed. During the earlier period, hyperventilation was the mainstay of our therapy, whereas during the later period, a more conservative approach (avoidance of hyperventilation) was adopted. In 1980 to 1981, 1 of the 6 patients (17%) who were eligible for ECMO by criteria of Bartlett et al survived, which is consistent with the published data. However, in 1986 to 1988, 9 of 10 ECMO-eligible patients (90%) survived (P less than .02). The corresponding survival figures using the alveolar-arterial oxygen difference criteria of Short et al were 0 of 5 survivors (0%) in 1980 to 1981 and 8 of 9 (89%) in 1986 to 1988 (P less than .006). These data indicate that approximately 90% of patients who are candidates for ECMO now survive in our institution without the use of that therapy.(ABSTRACT TRUNCATED AT 250 WORDS)